Pawel Lewicki

Paul Lewicki (born Pawel Lewicki; October 7, 1953) is a Polish-born American cognitive scientist, an entrepreneur, and investor. He was a professor of cognitive psychology at the University of Tulsa from 1984 through 2009, where he established the Nonconscious Information Processing Laboratory, funded by NSF and NIH. After leaving the university in 2009, he assumed the full-time role of CEO of StatSoft, a multinational analytics software company that he founded and where he was a majority shareholder. On March 24, 2014, StatSoft was sold to Dell.

Research in cognitive psychology

Lewicki is best known for his research on nonconscious information processing and self-perpetuation, where he demonstrated that procedural knowledge in many cases can be created via subconscious acquisition of information about complex covariations between events or features, and that subconsciously, individuals can acquire very complex knowledge structures based on highly-multidimensional patterns of data.

Lewicki's research on the self-perpetuating development of encoding dispositions purports to show that accidentally acquired (and even very slight) cognitive preferences or other encoding/interpretive dispositions can gradually develop and strengthen in a self-perpetuating manner. This mechanism may contribute to the development of [personality traits and individual preferences, it may facilitate learning, but it can also lead to self-perpetuating development of dysfunctional biases, phobias, aversions and other symptoms of personality disorders.

Cognitive mining versus predictive data mining

Lewicki together with fellow cognitive researcher Thomas Hill were among the first to publish evidence that advanced expertise acquired by humans via experience, involves the acquisition (learning) and utilization of repeatable patterns in data that are structurally more complex than what humans can verbalize or intuitively experience, because they involve high-order interactions between multiple variables. The human consciousness usually cannot handle more than only third-order interactions.

This approach became popular and rapidly adopted by the corporate world as so-called “predictive data mining” starting in the late 1990s and StatSoft’s STATISTICA Data Miner is now one of the widely used enterprise-level software systems for data mining and predictive analytics.

Lewicki began to write data analysis software for personal computers even before PCs became powerful enough to handle computationally-intensive procedures such as data mining.